web-dev-qa-db-ja.com

関数にパラメーターを渡す方法

現在、ビューを作成しています 回答のここを参照してください 。 crosstab()クエリに基づいて関数を作成して、日付を渡し、特定の日付のデータを取得するにはどうすればよいですか？

また、その関数を複数回呼び出して、異なる日付を渡してチャートを埋めるのもよい習慣ですか（たとえば）。

postgresqlviewfunctions
4
alexsmn

私はSQL関数を提案します：

_CREATE OR REPLACE FUNCTION foo(_date date)
  RETURNS TABLE (
   name       text  -- types have to match your actual types!
 , keyword_id int
 , project_id int
 , the_date   date
 , today      int
 , yesterday  int
 , week       int
 , month      int) AS
$func$

SELECT k.name, f.keyword_id, f.project_id, _date -- AS the_date -- col alias irrelevant
     , f.t AS today, f.y As yesterday, f.w AS week, f.m AS month
FROM   crosstab(

    -- crosstab function from previous question here
    -- http://dba.stackexchange.com/a/71266/3684

   ) f (rn int, keyword_id int, project_id int
       , t int, y int, w int, m int)
JOIN   keyword k USING (keyword_id);

$func$  LANGUAGE sql;
_

コール：

_SELECT * FROM foo('2014-07-07');
_

すべてのnow()::dateを__date_という名前の入力パラメーターに置き換えます。
Postgres 9.1以前では、SQL関数で位置パラメータ_$1_を使用します（どのバージョンでも使用できます）。

その他のコード例：

トリッキーなディテール

crosstab()関数は、クエリstringsをパラメーターとして受け取ります。関数パラメーターはcrosstab()内では表示されません。したがって、日付値をstring literalsとして渡す必要があります！
便宜上、関数format()をお勧めします。たとえば、2番目のパラメータは次のようになります。

_,format('VALUES(%L::date), (%L), (%L), (%L)'
        , $1, $1 - 1, $1 - 7, ($1 - interval '1 month')::date
       )
_

の代わりに：

_,$$
 VALUES
       (now()::date)
     , (now()::date - 1)
     , (now()::date - 7)
     , ((now() - interval '1 month')::date)
 $$
_

完全なコード

SQL関数

_CREATE OR REPLACE FUNCTION foo_sql(_date date)
  RETURNS TABLE (
   name       text
 , keyword_id int
 , project_id int
 , the_date   date
 , today      int
 , yesterday  int
 , week       int
 , month      int) AS
$func$
BEGIN

SELECT k.name, f.keyword_id, f.project_id, _date
     , f.t, f.y, f.w, f.m
FROM   crosstab (
   $$
   SELECT rn
        , pk.keyword_id
        , pk.project_id
        , d.created_at
        , COALESCE(pr.pos, 0)
   FROM  (SELECT *, row_number() OVER () AS rn FROM project_keyword) pk
   CROSS  JOIN (
       $$
    || format('VALUES(%L::date), (%L), (%L), (%L)'
              , $1, $1 - 1, $1 - 7, ($1 - interval '1 month')::date)
   || $$
     ) d(created_at)
   LEFT JOIN (
      SELECT keyword_id
           , project_id
           , created_at::date AS created_at
           , min(position) AS pos
      FROM   project_report
      GROUP  BY keyword_id, project_id, created_at::date
      ) pr USING (keyword_id, project_id, created_at)
   ORDER  BY pk.rn, d.created_at
   $$

  ,format('VALUES(%L::date), (%L), (%L), (%L)'
              , $1, $1 - 1, $1 - 7, ($1 - interval '1 month')::date)
   ) f (rn int, keyword_id int, project_id int
       , t int, y int, w int, m int)
JOIN   keyword k USING (keyword_id);

END
$func$  LANGUAGE sql;
_

PL/pgSQL関数

より短く、VALUES式を再利用します。おそらくもう少し速いです。

_CREATE OR REPLACE FUNCTION foo_plpgsql(_date date)
  RETURNS TABLE (name text, keyword_id int, project_id int, the_date date
               , today int, yesterday int, week int, month int) AS
$func$
DECLARE
   _dates text := format('VALUES(%L::date), (%L), (%L), (%L)'
                        , $1, $1 - 1, $1 - 7, ($1 - interval '1 month')::date);
BEGIN

SELECT k.name, f.keyword_id, f.project_id, _date, f.t, f.y, f.w, f.m
FROM   crosstab (
  'SELECT rn
        , pk.keyword_id
        , pk.project_id
        , d.created_at
        , COALESCE(pr.pos, 0)
   FROM  (SELECT *, row_number() OVER () AS rn FROM project_keyword) pk
   CROSS  JOIN (
       ' || _dates || '
     ) d(created_at)
   LEFT JOIN (
      SELECT keyword_id
           , project_id
           , created_at::date AS created_at
           , min(position) AS pos
      FROM   project_report
      GROUP  BY keyword_id, project_id, created_at::date
      ) pr USING (keyword_id, project_id, created_at)
   ORDER  BY pk.rn, d.created_at'

  ,_dates
   ) f (rn int, keyword_id int, project_id int, t int, y int, w int, m int)
JOIN   keyword k USING (keyword_id);

END
$func$  LANGUAGE plpgsql;
_

CTE

完全を期すために、関数を永続化せずに「同じ」を CTEs で指定します。

_WITH d(day)   AS (SELECT '2014-07-07'::date)  -- provide your date here
,    v(dates) AS (
   SELECT format('VALUES(%L::date), (%L), (%L), (%L)'
                 , day, day - 1, day - 7
                 ,(day - interval '1 month')::date)
   FROM d
   )
SELECT k.name, f.keyword_id, f.project_id, d.day AS the_date
     , f.t AS today, f.y As yesterday, f.w AS week, f.m AS month
FROM   crosstab (
  'SELECT rn
        , pk.keyword_id
        , pk.project_id
        , d.created_at
        , COALESCE(pr.pos, 0) AS pos
   FROM  (SELECT *, row_number() OVER () AS rn FROM project_keyword) pk
   CROSS  JOIN (
       ' || (SELECT dates FROM v) || '
     ) d(created_at)
   LEFT JOIN (
      SELECT keyword_id
           , project_id
           , created_at::date AS created_at
           , min(position) AS pos
      FROM   project_report
      GROUP  BY keyword_id, project_id, created_at::date
      ) pr USING (keyword_id, project_id, created_at)
   ORDER  BY pk.rn, d.created_at'

  ,(SELECT dates FROM v)
   ) f (rn int, keyword_id int, project_id int
       , t int, y int, w int, m int)
JOIN   keyword k USING (keyword_id);
_
10
Erwin Brandstetter